Does Amount of Protein in Formula Matter for Low-Birthweight Infants? A Cochrane Systematic Review 
From the * Faculty of Nursing and Correspondence: Shahirose Premji, RN, PhD, Assistant Professor and Neonatal Nurse Practitioner, University of Calgary, Faculty of Nursing, 2500 University Drive NW, Calgary, Alberta, Canada T2N 1N4. Electronic mail may be sent to premjis{at}ucalgary.ca.
Background: High protein intake may be associated with negative
consequences such as acidosis, uremia, and elevated levels of circulating
amino acids (eg, phenylalanine levels). We performed a systematic review of
randomized controlled trials to determine whether formula-fed low-birthweight
infants could tolerate protein intakes Protein is an important component of adequate nutrition because it provides essential amino acids required for protein synthesis, which is necessary for growth. Adequate amounts of protein are also required for bone and blood constituents, turnover of tissues, synthesis of hormones and enzymes, and maintenance of oncotic pressure.1 Deficiency of protein in infants leads to growth failure and, when extreme, can lead to lower resistance to infection.2 For premature infants, a range of advisable protein intake is determined by combining information given by the factorial, metabolic, and neurodevelopmental approaches.3 The enteral protein intake required for preterm infant growth and protein accretion according to the factorial approach using the "reference fetus"4,5 is estimated to be 3.5–4.0 g/kg/d for premature infants.6 If energy intakes are maintained at the recommended range7 with formulas currently available for preterm infants in North America, which contain 3 g of protein per 100 kcal, formula-fed infants would receive about 3.2–4.1 g/kg/d of protein. However, premature infants may not be able to handle higher protein intakes efficiently due to immaturity of amino acid metabolic pathways.3 Increased concentrations of amino acids such as tyrosine and phenylalanine concentrations, hydrogen ions, and urea may result.3 Theoretically, metabolic changes could be harmful in terms of mental development.
The balance between supposed benefits and risks of higher protein intake
for formula-fed low-birthweight infants <2.5 kg remains unclear. We
therefore performed a systematic review of the literature to determine whether
high (
Searching We identified randomized controlled trials comparing 3 categories of protein intake—low (<3.0 g/kg/d), high (  3.0 g/kg/d but <4.0
g/kg/d), or very high (>4.0 g/kg/d)—in infants who weighed <2.5 kg
at birth, whether appropriate or small for gestational age (AGA or SGA), and
were studied during their initial hospital stay. Two reviewers (S.P. and T.F.)
independently identified trials from MEDLINE (1966 to May 2005), CINAHL (1982
to May 2005), PubMed (1966 to May 2005), EMBASE (1980 to May 2005), and the
Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library,
Issue 2, 2005), abstracts, conferences and symposia proceedings from Society
of Pediatric Research, and American Academy of Pediatrics. MeSH headings
including infant, new-born, low birth weight, small for gestational age,
very low birth weight, premature, amino acids, dietary proteins, milk
proteins, milk, infant food, food, and infant formula and text
words including formula and protein were used for the
computerized searches. References were reviewed independently for additional
relevant titles and abstracts for articles up to 50 years old. No language
restrictions were applied.
Study Selection and Validity Assessment
Standard methods of the Cochrane Collaboration and of the Cochrane Neonatal Review Group were used to independently assess (unblinded) trial quality. Trials were rated yes, no, or don't know on the following criteria: blinding of randomization, blinding of intervention, complete follow-up, and blinding of outcome measurement. Data for this review were extracted independently (S.P. and T.F.). Efforts were made to contact investigators for data, additional information, or clarification regarding studies. Differences at any stages of the review process were resolved, with the third reviewer (R.S.) facilitating consensus decision making. A standardized statistical method was used to handle 3-arm trials where 2 groups fell within 1 predesignated protein intake group.12 Meta-analysis was performed using the software of the Cochrane Review Manager 4.1 (Cochrane Collaboration, Nordic Cochrane Centre, Denmark), and a fixed-effect model was used throughout the review. To estimate the effect of an intervention, weighted mean differences (WMD) were reported for continuous variables, and typical estimates for relative risk and risk difference were reported for categorical variables. Uncertainty in each result was expressed as 95% confidence intervals (CI). A statistical test for heterogeneity (I2 test) included in the graphical output of Cochrane Reviews was used to assess variability in treatment effects being evaluated in the different trials.
Eligible Studies for Inclusion We identified 53 studies in the literature search, 37 of which we reviewed in detail, and 5 studies were included in this review (see Figure 1).13–17 Multiple reports of the same study were considered single trials. Schulze et al18 published data on energy expenditure and energy balance, which were collected in a subset of infants in the Kashyap et al15 study. Svenningsen et al reported short-term outcomes16 and long-term follow-up growth parameters and neurodevelopmental outcomes up to 2 years of age19 in separate papers.
Study Characteristics
In terms of methodologic quality of trials, infants were allocated to assigned treatment by randomization in all studies included in this review. Only 2 studies13,14 reported adequate concealment of allocation and blinding of randomization. Three studies13–15 reported that the intervention was blinded to caregivers or investigator(s). One study13 reported blinding of outcome. Only 1 study reported an intention-to-treat analysis.17
Meta-analysis Data on weight gain favors the high protein group (WMD 2.36 g/kg/d, 95% CI, 1.31–3.40; see Figure 2). Given the significant heterogeneity of treatment effect (I2 = 56.7%), the data need to be interpreted with care. Findings were not consistent across studies, with Bhatia et al13 and Svenningsen et al16 reporting no significant differences and Hillman et al,14 Kashyap et al,15 and Wauben et al17 reporting significantly greater weight gain in infants receiving high protein intakes. For the outcome of linear growth (see Figure 2), the overall analysis revealed no significant difference between groups (WMD, 0.16 cm/wk; 95% CI, –0.02 to 0.34). Although 4 studies13–15,16 compared head growth in infants receiving high vs low protein intakes, data were missing for 3 studies13,14,16; hence, a meta-analysis was not feasible for this review. Kashyap et al15 found that infants receiving high protein intakes had significantly greater head growth. Bhatia et al,13 Hillman et al,14 and Svenningsen et al16 reported no significant difference in head growth. Blood urea nitrogen levels measured at the 2-week point were significantly higher in infants in the high protein intakes group (WMD, 1.92 mg/dL; 95% CI, 1.00–2.84; see Figure 3) compared to the low protein group. The study by Svenningsen et al16 measured blood urea nitrogen levels at different time points than the other studies and, as a result, was not included in the meta-analysis. Svenningsen et al16 did not find a significant difference in blood urea nitrogen at the third and fifth week of life, although at 7 weeks, levels were significantly higher among the infants receiving higher protein intakes (third-week p = .85, fifth-week p = .375, and seventh-week p = .0005).
Meta-analysis for nitrogen accretion revealed significantly higher nitrogen accretion (WMD, 143.7 mg/kg/d; 95% CI, 128.7–158.8; see Figure 4) in infants receiving formula with high protein content compared with infants receiving the low-protein formula. Given the significant heterogeneity of treatment effect (I2 = 91.5%), the data need to be interpreted prudently. Both studies15,17 included in the meta-analysis reported statistically significant higher protein accretion in the high-protein-formula groups.
In this systematic review, the categorical outcome necrotizing enterocolitis was defined as Bell's Stage II or greater. The meta-analysis showed no significant effect of protein intake on necrotizing enterocolitis (typical risk difference, 0.00; 95% CI, –0.12 to 0.12; see Figure 5). However, it is not clear what operational definition was used in the 2 studies16,17 which contributed data to this meta-analysis and reported no incidence of necrotizing enterocolitis in either the high- or the low-protein-intake groups.
This review determined that there were benefits in weight gain and nitrogen accretion without any clear risks associated with high protein intakes ( 3.0 g/kg/d but <4.0 g/kg/d). One needs to be prudent in interpreting
the findings of this systematic review, given the small number of trials that
met the inclusion criteria, methodologic limitations of included trials, and
absence of relevant data on long-term effects and potential adverse effects of
high protein intake in formulafed "healthy" preterm infants.
Thirty-two of the 37 studies were excluded because they did not compare
sufficiently different protein intakes or they examined a different
intervention (eg, studies examining quality of protein). The validity of the
analysis of the 5 included trials is threatened as a result of limitations
such as lack of adequate concealment allocation, clinical variability (eg,
variability in participants, interventions, and outcomes assessed), and
methodologic variability (eg, differences in trial design, duration of
intervention, or study period). Of these limitations, one of particular
importance is the small difference (range, 0.56–1.36 g/kg/d) in protein
intake among comparison groups in some of the individual trials. All of the
abovementioned limitations may be potential sources of statistical
heterogeneity in measures of weight gain and nitrogen accretion and may
explain the differences in treatment effects.
Weight gain (g/kg/d) was the most commonly reported outcome among the 5 studies included in this review.13–17 There was an overall increase in weight gain in infants randomized to the high-protein-intake group compared with the low-protein-intake group (WMD, 2.36 g/kg/d; 95% CI, 1.31–3.40). Weight gain is an important variable in growth assessment but a nonspecific indicator of the adequacy of protein intake as it provides no information about the changes in body composition.3 There was significantly greater nitrogen accretion (WMD, 143.7 mg/kg/d; 95% CI, 128.7–158.8) in infants randomized to the high-protein-intake groups, which suggests that some or all of the increment in weight is due to gains in lean body mass. These findings indicate that higher protein intakes may help correct the nonoptimal body composition seen in preterm infants at term-adjusted age.20 However, the ideal composition of weight gain is not known in preterm infants, though it is suggested that the lower lean tissue and higher fat gain of these infants relative to the fetus may not be desirable.18
Infants with very low birthweights and extreme prematurity may not be able
to tolerate high ( Studies by Goldman et al21 and Raiha et al,24 which assessed protein intakes above 4.0 g/kg/d, could not be included in this systematic review, because the higher protein formula differed significantly (>10%) in relative concentration of electrolytes and minerals, hence violating one of the inclusion criteria. The quantities of higher protein intake in these studies were 4.5 g/kg/d21 and 6–7.2 g/kg/d.24 Multiple reports were published for the Goldman et al21–23 and Raiha et al,24,25 Gaull et al,26 and Rassin et al27 studies. Goldman et al,22 who administered the Stanford-Binet test at 3 and 5–7 years of age, noted a significant increase in the incidence of low IQs among infants with birthweights <1300 g who were fed very high protein intakes of 6–7.2 g/kg/d during their initial hospitalization. The findings of the above studies cannot be attributed directly to the high-protein intakes because other nutrients were not held constant when evaluating the protein intake. Neurodevelopmental outcomes of early nutrition (other than IQ score and Bayley score at 18 months or later, which were of primary interest for this systematic review) were evaluated by 2 studies13,19 included in this systematic review. Bhatia et al13 used the Neonatal Behavioral Assessment scale that has known psychometric properties but has been validated for use only in term infants up to 2 months of life.28 The Bhatia et al13 results suggested improvements with high protein intakes compared with low protein intakes in some of the parameters of neurodevelopmental outcome, such as orientation, habituation, autonomic stability. Svenningsen et al19 and reported no significant differences in neurodevelopmental outcomes up to 2 years of age. They assessed developmental performance indicators such as sitting, standing, walking, and talking at 5–6, 10–11, 14–18, and 24 months of age on 46 of the 48 infants enrolled in the study. At 10–14 months, they also performed an audiometric test. The instruments used for these assessments were not stated. Other potential adverse effects of high protein intake were assessed by examining days to full feedings, necrotizing enterocolitis, sepsis, and diarrhea. However, limited information could be obtained from the included studies regarding these potential risks. Although a meta-analysis was carried out for necrotizing enterocolitis, the findings presented should be interpreted cautiously because of (1) the uncertainty about the definition of necrotizing enterocolitis used by some studies and (2) the small number of infants in the 2 groups (n = 49 receiving high protein intake and n = 38 receiving low protein intake).
This systematic review suggests that weight gain and nitrogen accretion can
be promoted by regulating protein intake in "healthy" formula-fed
preterm The American Academy of
Pediatrics6 infants.
and the Canadian Pediatric
Society7 recommend
3.0–4.0 g/kg/d of protein for preterm infants. In "healthy"
formula-fed low-birthweight infants, accelerated weight and nitrogen accretion
was noted with higher protein intakes ( This paper is a version of a Cochrane systematic review (Premji S, Fenton T, Sauve R. Higher vs lower protein intake in formula-fed low birth weight infants) that was prepared under the aegis of the Cochrane Collaboration and published in the Cochrane Library, Issue 1, 2006 (see http://www.thecochranelibrary.com for information). Cochrane systematic reviews are regularly updated as new research becomes available and in response to feedback from readers. Please consult the Cochrane Library for the most recent version of this review. The Cochrane Collaboration's publication policy permits journals to publish reviews, with priority if required, but permits the Cochrane Collaboration also to publish and disseminate such reviews. We thank Dr Rollin Brant for statistical advice. This systematic review was supported by the Perinatal Research Fund provided by PREMI (Partnership for Research and Education for Mothers and Infants). Received for publication April 26, 2006. Accepted for publication July 26, 2006.
Journal of Parenteral and Enteral Nutrition, Vol. 30, No. 6,
507-514 (2006)
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Top
MATERIALS AND METHODS
RESULTS
